US8495523B2ActiveUtilityA1
Gradient-based search mechanism for optimizing photolithograph masks
Est. expiryAug 5, 2029(~3.1 yrs left)· nominal 20-yr term from priority
G03F 1/36
57
PatentIndex Score
0
Cited by
20
References
3
Claims
Abstract
A mechanism is provided for optimizing a photolithograph mask. A given target pattern is received. An initial fictitious mask is generated from the given target pattern and an initial value of α 2 is selected where the initial value of α 2 is used to determine a light intensity and a wafer image. The light intensity for each pixel in the initial fictitious mask and the wafer image for each pixel in the initial fictitious mask are then determined. A determination is then made as to whether a convergence has been achieved by comparing the wafer image generated from the fictitious mask to the given target pattern. Responsive to a convergence of the wafer image generated from the fictitious mask to the given target pattern, a final mask is generated to use to transfer an image to a wafer.
Claims
exact text as granted — not AI-modifiedWhat is claimed is
1. A method, in a data processing system, for optimizing a photolithograph mask, the method comprising:
receiving, by the data processing system, a given target pattern;
generating, by the data processing system, an initial fictitious mask;
selecting, by the data processing, system, an initial value of α 2 , wherein the initial value of α 2 is used to determine a light intensity and a wafer image;
determining, by the data processing system, the light intensity for each pixel in the initial fictitious mask, wherein the light intensity is determined using the following equation:
I
j
=
∑
k
=
1
m
σ
k
∑
i
=
1
MN
ϕ
ijk
S
2
(
θ
i
)
2
wherein I j is the light intensity at pixel location j, wherein θ is a specific pixel, wherein m is an eigenvector of a kernel, wherein M is a number of grids in a x direction, wherein N is a number of grids in a y direction, wherein S 2 is a continuous first-order derivative function, wherein σ is an eigenvalue of an optical kernel function, wherein φ is eigenvector of the optical kernel function, wherein k is a kernel index, and wherein i is a grid index;
determining, by the data processing system, the wafer image for each pixel in the initial fictitious mask;
determining, by the data processing system, whether a convergence has been achieved by comparing the wafer image generated from the initial fictitious mask to the given target pattern;
responsive to a convergence of the wafer image generated from the initial fictitious mask to the given target pattern, generating, by the data processing system, a final mask to use to transfer an image to a wafer;
responsive to a failure of the wafer image generated from the initial fictitious mask to converge with the given target pattern, determining, by the data processing system, a gradient for each pixel in the initial fictitious mask;
updating, by the data processing system, the initial fictitious mask using the gradient thereby forming an updated fictitious mask; and
repeating, by the data processing system, the steps of determining the light intensity for each pixel in the updated fictitious mask, determining the wafer image for each pixel in the updated fictitious mask, and determining whether the convergence has been achieved by comparing the wafer image generated from the updated fictitious mask to the given target pattern until the wafer image generated from the initial fictitious mask converges with the given target pattern.
2. A computer program product comprising a non-transitory computer recordable medium having computer readable program recorded thereon, wherein the computer readable program, when executed on a computing device, causes the computing device to rating device to:
receive a given target pattern;
generating an initial fictitious mask;
select an initial value of α 2 , wherein the initial value of α 2 is used to determine a light intensity and a water image;
determine the light intensity for each pixel in the initial fictitious mask, wherein the computer readable program to determine the light intensity uses the following equation:
I
j
=
∑
k
=
1
m
σ
k
∑
i
=
1
MN
ϕ
ijk
S
2
(
θ
i
)
2
wherein I j is the light at intensity at pixel location j, wherein θ is a specific pixel, wherein m is an eigenvector of a kernel, wherein M is a number of grids in a x direction, wherein N is a number of grids in a y direction, wherein S 2 is a continuous first-order derivative function, wherein σ is an eigenvalue of an optical kernel function, wherein φ is eigenvector of the optical kernel function, wherein k is the a kernel index, and wherein i is a grid index;
determine the wafer image for each pixel in the initial fictitious mask;
determine whether a convergence has been achieved by comparing the wafer image generated from the initial fictitious mask to the given target pattern;
responsive to a convergence of the wafer image generated from the initial fictitious mask to the given target pattern, generate a final mask to use to transfer an image to a wafer;
responsive to a failure of the wafer image generated from the initial fictitious mask to converge with the given target pattern, determine a gradient for each pixel in the initial fictitious mask;
update the initial fictitious mask using the gradient thereby forming an updated fictitious mask; and
repeat the steps of determining the light intensity for each pixel in the updated fictitious mask, determining the wafer image for each pixel in the updated fictitious mask, and determining whether the convergence has been achieved by comparing the wafer image generated from the updated fictitious mask to the given target pattern until the wafer image generated from the initial fictitious mask converges with the given target pattern.
3. An apparatus, comprising:
a processor; and
a memory coupled to the processor, wherein the memory comprises instructions which, when executed by the processor, cause the processor to:
receive an initial photolithograph mask and a given target pattern;
generate an initial fictitious mask;
select an initial value of α 2 wherein the initial value of α 2 is used to determine a light intensity and a wafer image;
determine the light intensity for each pixel in the initial fictitious mask, wherein the istructions to determine the light intensity uses the following equation:
I
j
=
∑
k
=
1
m
σ
k
∑
i
=
1
MN
ϕ
ijk
S
2
(
θ
i
)
2
wherein I j , is the light intensity at pixel location j, wherein θ is a specific pixel, wherein m is an eigenvector of a kernel, wherein M is a number of grids in a x direction, wherein N is a number of grids in ay direction, wherein S 2 is a continuous first-order derivative function, wherein σ is an eigenvalue of an optical kernel function, wherein φ is eigenvector of the optical kernel function, wherein k is a kernel index, and wherein i is a grid index;
determine the wafer image for each pixel in the initial fictitious mask;
determine whether a convergence has been achieved by comparing the wafer image generated from the initial fictitious mask to the given target pattern;
responsive to a convergence of the wafer image generated from the initial fictitious mask to the given target pattern, generate a final mask to use to transfer an image to a wafer;
responsive to a failure of the wafer image generated from the initial fictitious mask to converge with the given target pattern, determine a gradient for each pixel in the initial fictitious mask:
update the initial fictitious mask using the gradient thereby forming an updated fictitious mask; and
repeat the steps of determining, the light intensity for each pixel in the updated fictitious mask, determining the wafer image for each pixel in the updated fictitious mask, and determining whether the convergences has been achieved by comparing the wafer image generated from the updated fictitious mask to the given target pattern until the wafer image generated from the initial fictitious mask converges with the given target pattern.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.